JevAny-Robotics-4B-LoRA / eval /robopulsepp.md
HuanxinSheng's picture
robojudge-4b-v02 SFT final (step 5780)
fac5b68 verified
|
Raw History Blame Contribute Delete
1.43 kB

robopulsepp-v1: 7720 items

model answered acc (95% CI) balanced acc macro-F1 F1 pos / neg AUROC NLL ECE pred pos rate
rj4b-v02-final 7720/7720 60.4 [57.9, 62.9] 65.9 55.8 70.0 / 41.6 0.728 0.649 0.095 51.0
JevAny-4B 7720/7720 69.9 [67.7, 72.0] 59.1 57.5 80.5 / 34.4 0.641 0.580 0.050 73.0

Timing

model GPUs load s wall s (loop) mean / median / p95 ms per item items/s median input tokens GPU-hours
rj4b-v02-final 1x NVIDIA A100-SXM4-40GB 4 2022 262 / 207 / 520 3.82 790 0.56
JevAny-4B 1x NVIDIA A100-SXM4-40GB 7 2034 263 / 210 / 515 3.80 790 0.57

Interval level (majority of kept pairs)

model intervals acc (95% CI) macro-F1
rj4b-v02-final 1062 67.6 [64.6, 70.4] 65.6
JevAny-4B 1062 67.3 [64.8, 70.0] 58.7

Slices (accuracy / macro-F1, %)

slice n rj4b-v02-final JevAny-4B
n_views=1 5097 65.9 / 60.2 70.0 / 58.6
n_views=2 1693 45.4 / 44.8 66.6 / 52.3
n_views=3 930 57.2 / 49.3 75.5 / 59.9

Paired: rj4b-v02-final − JevAny-4B

  • accuracy: -9.6 pts, 95% CI [-12.6, -6.2]
  • macro-F1: -1.7 pts, 95% CI [-4.8, 1.8]
  • McNemar: rj4b-v02-final only correct 995, JevAny-4B only correct 1733, exact p = 9.384e-46